WO1997001756A1

WO1997001756A1 - Estimating grain size in geological samples

Info

Publication number: WO1997001756A1
Application number: PCT/GB1995/002424
Authority: WO
Inventors: Colin Leonard Bird; Sydney George Chapman
Original assignee: International Business Machines Corporation
Priority date: 1995-06-29
Filing date: 1995-10-13
Publication date: 1997-01-16
Also published as: EP0778943A1; GB9513300D0; GB2302736A

Abstract

A method and system for estimating the grain size in geological samples such as cores is disclosed. This technique is achieved by a data processing system wherein data of a geological sample is retrieved from a storage device, the storage device storing data of the geological sample which incorporates at least one characteristic varying with the grain size of the sample. Responsive to a selection of a portion of the data, a subset of the data corresponding to the selected portion is retrieved, the subset of data incorporating said at least one characteristic varying with the grain size. Next, a set of variations of said at least one characteristic in said subset of the data is evaluated in order to determine an estimate of the average grain size of the portion with respect to a dominant variation of said set. Further, in preferred embodiments, a digital image of the geological sample representative of said data is displayed on a display device in order to allow a user to select the portion of the image to be evaluated. Then the determined estimate is made available to the user.

Description

ESTIMATING GRAIN SIZE IN GEOLOGICAL SAMPLES

The invention relates to a system and method for estimating grain size in a geological formation, in particular in geological samples such as cores.

Typically a company involved in oil or gas exploration extracts a number of exploration cores using techniques such as percussion or rotary drilling. These cores are subsequently examined by geologists to understand the way in which the formation was laid down and has since evolved. This examination allows the geologists to evaluate the most likely locations where oil or gas can be found.

In order to maintain the integrity of the information carried in the core it is important to the geologist that a clear record of what is observed is kept. The aim of a log is to act as an informative but succinct summary of the core, rather than a geological essay, it is useful policy to write the log description systematically, in a set order, to ensure that no important information is omitted. The range of rock properties recorded in writing and the order in which they are set down will depend on the lithologies cored and the nature of the work being undertaken. For example, the order suggest by the American Association of Petroleum Geologists for sedimentary rocks is as follows: 1) Rock type - followed by classification; 2) Colour;

3) Grain size, roundness and sorting, etc.;

4) Cement and/or matrix materials;

5) Fossils and accessories;

6) Sedimentaries structures; 7) Porosity and oil shows.

From the above list it is clear that a very useful piece of information which can be found from such cores is the grain size of the sediments. Moreover, the grain size is relevant not only for petroleum investigation but also for engineering geologists, for whom mechanical properties are more important than mineralogy and texture, as described in Knill et al. , The logging of rock cores for engineering purpose, Q.J. Eng. Geol. 3, pp. 1-24.

However, due to their considerable weight and size (a core can be more than 100 . long) , such cores are moved and stored with difficulty, and cannot be distributed in their entirety to more than one place at once. Therefore, a geologist wishing to perform a direct examination of the geological sample, is often required to collect some data from the locations where the cores are stored and then to return to the laboratory for the final analysis. Moreover, if they are kept permanently at all, it is generally in some remote core stores where warehousing is cheap, and visits are therefore made only in cases of special needs.

Moreover to facilitate easier handling and storing, the core is often cut to form a set of different portions; for instance it may be split into 1 m. boxes. In some stores there is only space to view one box at a time, causing obvious difficulties during the examination of the cores. In addition portions of the core can be easily lost or replaced in a different order jeopardizing the results of subsequent examinations. Eventually, the core samples inevitably deteriorate with age, some more than others, either naturally or by excessive handling and sampling.

Therefore the gap between the information which can be derived from a detailed log, and that obtainable from the original core, can largely be filled with a good set of core photographs, such as the one shown in Figure 1 (reproduced from G.A. Blackbourn, Cores and Core Logging for geologists, Whittles Publishing, 1990, p. 74) , which can be copied, are easily moved, can be readily maintained in a proper order, and do not deteriorate significantly with age.

However, up until now, a direct human examination, i.e. checking by eye, of the image of a core, rather than the core itself, is still required by the geologists in order to detect important characteristics of the core such as the grain size. Obviously this is very time consuming, labour intensive, and prone to errors, and so it would be highly desirable to reduce the time that a geologist must spend inspecting the core or its image.

It is hence an object of the present invention to provide an apparatus and method for estimating the grain size in the digital image of a geological sample, which will alleviate the necessity for direct human estimation.

Accordingly the present invention provides a method of operating a data processing system to estimate grain size in a geological sample, comprising the steps of: retrieving from a storage device data of the geological sample, the data storage device storing data of the geological sample which incorporates at least one characteristic varying with the grain size; responsive to a selection of a portion of the data, retrieving a subset of the data corresponding to the selected portion, the subset of data incorporating said at least one characteristic varying with the grain size; evaluating a set of variations of said at least one characteristic in said subset of the data; and determining an estimate of the average grain size of the portion with respect to a dominant variation of said set.

The first step in achieving this is to store images of exploration cores in an on-line digital form, in order to avoid not only the moving of the cores or of the geologists between the storing locations and the laboratories but also of the photographs themselves.

Further, storing digital images into an electronic database for subsequent examination makes available to the geologists a larger amount of data which might be helpful for their evaluations. In fact images of cores stored in distant locations such as in other countries can be used for comparison purposes with the images to be examined.

Moreover, digital images can be very useful when explorations are performed in remote locations, eg in off-shore drilling, where most likely a quick and accurate examination is required to reduce the incidence of costly and useless drilling activities. Thus, images are captured as the core is extracted and then transmitted to a remote office for an easier and more complete inspection.

In addition, a large amount of data associated with each image of the core can be retrieved. According to the invention, in order to assist the geologist in examination of such data, a set of portions of the image having similar features can be identified by producing the estimation of the grain size automatically rather than under the guidance of a human user; this might be achieved by a system for automatically selecting in the whole image the portions with a specified trend of grain size. Indeed, the process could be used to identify similar trends in several cores, rather than just a single core.

Typically a user, such as a geologist, will be involved in the process, since the grain size trend in the exploration cores is just one element to be determined before evaluating the likelihood of the existence of a gas or oil fields in the survey location. Accordingly, in preferred embodiments, the step of retrieving from the storage device data of the geological sample comprises retrieving data necessary to display a digital image of the geological sample representative of said data, and displaying such an image on a display device. Further, the step of retrieving a subset of the data corresponding to the portion is responsive to a user selection of a portion of the image, and the step of determining an estimate of the average grain size includes the step of making available to the user such an estimate. Preferably, the dominant variation is identified by applying a Fourier Transform to said subset of the data so to generate a new representation of said subset based on spatial variations. The use of Fourier transforms in image processing is well known and so will not be described in detail herein; the book by Rhys Lewis entitled "Practical

Digital Image Processing", Ellis Horwood (1990) pp. 150-161 describes this use in detail.

In a particular embodiment the estimate of the average of the grain size is determined by dividing the dominant variation in the at least one characteristic by the resolution of the data.

In addition, in order to simplify the user selection of the portion of the image, in an advantageous embodiment the user selection of the portion of the image is limited to the selection of at least one selecting parameter, the others being preset. Further, the preset selecting parameters include the angle and the dimension of the portion and the at least one parameter selected by the user includes the location of the portion.

Moreover, in order to reduce the complexity of the calculation, consequently providing a faster estimation of the grain size, the selection of a portion of the image is, in preferred embodiments, limited to a selection of a line thereof only.

In the image of a piece of core, whether digital or photographic, the recorded amplitude is a measure of the light intensity reflected from the surface of the core. The amplitude is expected to vary with the profile of the individual grains within the core surface, given that adequate resolution is used. Accordingly, in a particular preferred embodiment the at least one characteristic is light intensity.

Moreover the user can be assisted in the choice of the portion by other characteristics of the image displayed which can locate interesting areas to be evaluated, such as colour or texture showing two regions to be different, or by an image processing operation that accentuates the structure of the image. Accordingly, in a preferred arrangement, the displayed image also incorporates at least a second characteristic of the geological sample which provides additional information to the user to facilitate the selection of the portion to be evaluated. Preferably the at least second characteristic is colour.

Typically, in a particular preferred arrangement the step of making available to the user such an estimate comprises the step of displaying a slider bar indicating the estimate, the bar being annotated with a set of different grades of grain size.

Viewed from a second aspect the present invention provides a data processing system for estimating grain size in a geological sample, comprising: a retrieval means for retrieving data of the geological sample from a storage device, the storage device storing data of the geological sample which incorporates at least one characteristic varying with the grain size; selection means for selecting a portion of the data retrieved; the retrieval means being responsive to the selection means to retrieve a subset of the data corresponding to the selected portion, the subset of data incorporating the at least one characteristic varying with the grain size; evaluating means for evaluating a set of variations of said at least one characteristic in said subset of the data; and estimating means for determining an estimate of the average grain size of the portion with respect to a dominant variation in said set.

In preferred embodiments, the system comprises a display means for producing a digital image of the geological sample representative of the data retrieved by the retrieval means and for displaying the digital image on a display device connectable to the system, and the selection means being responsive to a signal received from an input device connectable to the system, to enable a user to select a portion of the displayed image.

The present invention solves the problem of estimating the grain size in a geological sample by providing an apparatus and a method which uses a number of image processing techniques to quickly and reliably estimate the grain size.

The present invention will be described further, by way of example only with reference to an embodiment thereof as illustrated in the accompanying drawings, in which:

Figure 1 is a copy of a photograph representing a core sample;

Figure 2 shows a representative data processing system;

Figure 3 is a flow diagram illustrating the process steps according to the preferred embodiment of the present invention;

Figure 4A shows an image an of exploration core;

Figure 4B shows an enlarged view of a section of the image in Figure 4A; Figure 4C is an intensity vector of a segment shown by the line 310 in Figure 4B;

Figure 5 is a graph illustrating a set of frequencies generated by applying the Fourier transform to the intensity vector as shown in Figure 4C;

Figure 6 depicts a core log.

Figure 2 illustrates a preferred embodiment of a computerized data processing system operating in conjunction with the technique illustrated in Figure 3 in order to perform the herein described operations of the embodiment. The system comprises a CPU 10, read only memory (ROM) 16, random access memory (RAM) 14, I/O adapter 18, user interface adapter 22, communications adapter 34, and display adapter 36, all interconnected via a common address/data/and control path or bus 12. Each of the above components accesses the common bus utilizing conventional techniques known to those of ordinary skill in the art, and includes such methods as dedicating particular address ranges to each component in the system, with the CPU being the busmaster. Other conventional techniques known to those of ordinary skill in the art employed in the system of Figure 1 include direct memory access (DMA) used to transfer data at high speed from external devices such as DASD 20 or the network 40 shown to the data processing system's RAM 14.

As is further shown in Figure 2, these external devices such as DASD 20 interface to the common bus 12 through respective adapters such as I/O adapter 18. Other external devices, such as the display 38 similarly use their respective adapter such as display adapter 36 to provide data flow between the bus 12 and the display 38 or other device. Various user interface means are provided for interconnection and use with the user interface adapter 22 which, in the figure, has attached thereto representative user input devices such as a joy stick 32, mouse pointer 26, and keyboard 24. Each of these units is well known as such and accordingly will not be described in detail herein.

As will hereinafter be detailed, upon implementation of an appropriate technique such as that described herein with reference to Figure 3, the system of Figure 2 will execute the method in order to estimate grain size in geological samples.

Passing now to Figure 3 the overall process will be described in detail. At the highest level, there are five stages 200, 210, 220, 230, 240 to the process. In step 200 a user by means of the data processing system of Figure 2 retrieves digital data corresponding to an image of a section of a core stored in the directly attached DASD 20 or in different memory devices accessible through the network 40. The images stored in the memory devices can be arranged as single image data files or in one or more databases to make the search of the images easier. The retrieved data is displayed as an image on the display screen 38. Then the control passes to step 210, wherein the user, who is viewing the image displayed on the display 38, identifies a portion of the image as being of interest, using a selection device, most likely the mouse pointer 26. The user can choose the portions randomly, or depending on the result of a previous estimation of another portion, or by other characteristics of the image displayed which can locate interesting areas to be evaluated, such as colour or texture showing two region to be different, or by an image process operation that accentuates the structure of the image, eg zooming.

This portion might be a line, which may be at any angle, not necessarily horizontal or vertical; the line is the 1-dimensional case. Alternatively the portion might be a rectangular area: the 2-dimensional case.

For both the 1- and the 2-dimensional case, techniques can be envisaged which would simplify the selection by taking advantage of parameters which could be preset by the user. For example, a length and an angle could be preset for a line portion, enabling selection to be made by a single mouse click.

The result of the selection stage is a set of values which precisely describe the location within the original data of the portion of interest.

In step 220, using the location values from the selection stage, the data values comprising the portion of interest are retrieved from the full set of data for the section of core, as stored in the memory device 20 or in the network 40. In the 1-dimensional (1-D) case, all points lying along the line are obtained; where a 2-dimensional (2-D) area has been selected, all points lying within the rectangle are extracted. Although subsequent processing is simpler for rectangular areas, other shapes are not necessarily precluded by the use of the term "rectangle".

The extracted values are a sample of the data for the portion of interest, and encode some characteristic which is representative of the grain size of the material in the original core section. Typically, this will be an intensity value recorded by a sensor which responds to light. However data values from a sonar sensor would be another example. Experience has shown that under perpendicular illumination, the intensity profile recorded in the image is a function of both the sphere diameter and the resolution at which the image is captured. This will result in a periodic variation in the intensity profile and the average period will yield a fair estimate of the grain size. Therefore, using the example of intensity, the sample will comprise either a vector (1-D) or a matrix (2-D) of intensity values.

In step 230, the estimation process starts. This stage is considered in greater detail below, as it is the key part of the whole process. In summary, a Fourier transform is applied to the sample, then the dominant frequency is obtained from the transform output. The period of this frequency, together with the resolution at which the original data was obtained, can be used to compute a value for the grain size.

Finally, in step 240, the computed grain size value can be shown to the user as a numeric value, but interaction with the process is better served if a more visual method is used. For example, a slider bar may be set to correspond to the computed grain size, where the bar is annotated with grades, such as: "Very Fine"; "Fine"; "Medium"; and "Coarse". Colour coding could also be employed. Other methods can also be envisaged.

Interaction is achieved by moving the selection position, and so getting rapid feed-back of the grain size estimate. This enables the user - the geologist - to gain an impression of both the distribution of sizes and the manner in which size varies from one part of the image to another.

Referring now to Figure 4A, an image 300 of a section of a core as displayed on display 38 to a user of the embodiment is shown. In Figure 4B an enlarged display of a part of the image 300 is shown, with a selected portion 310. In Figure 4C, the graph represents the light intensity of each pixel corresponding to the selected portion 310. Each intensity value then is stored in the intensity vector, or matrix defining the Pixel Line Intensity Profile of the selected portion.

Returning back to step 230, which dealt with estimating the grain size, different ways to implement the Fourier transform are well known to those skilled in the art so need not be described any further herein. The output of the transform will usually have both real and imaginary parts, from which a magnitude value can be obtained. Magnitude values are used in the estimation.

The Fourier transform can be used to generate a new representation of the image data, based on spatial frequencies, while preserving the information in the original image. Each frequency will have its own magnitude, indicative of how much it contributes to the overall image. In turn, the frequencies indicate the periods with which the main spatial characteristics recur. Accordingly, if the grains of the sediments in the core are considered as small balls packed together, the diameter of the ball represents a spatial period.

For simplicity, the following description assumes the 1-D case, with reference to Figure 5, which is reproduced from G.A. Blackbourn, Cores and Core Logging for geologists, Whittles Publishing, 1990, p. 63. However it is clear that the embodiment is not intended to be limited in any way to such a particular case. For instance, a 2-D case can be easily deduced as a sum of a set of 1-D cases.

There will always be a fundamental frequency corresponding to the full length of the vector, this being the lowest frequency that can be fitted into that space. After that, the next higher frequency is that for which the period can be fitted twice into the length of the vector. After that comes the period for three times, then four times, and so on. The limiting case comes with a period equal to one vector sample. By examining the magnitude of each of these contributing frequencies, it is clear which is the dominant one (other than the fundamental) .

The period is a spatial measure which, from the ball model used above, corresponds to the diameter of a grain. The direct relationship comes from the sampling interval of the original data, referred to earlier as the resolution. The two values together give an estimate of the grain size. With reference to the Figure 5 example in which the amplitudes have been chosen merely for the purpose of illustration, consider a line 500 containing 32 samples (which will appear on the display screen as 32 pixels) . It can be seen that the dominant frequency occurs when four cycles are fitted into the length of the line, then each cycle comprises 8 samples (or pixels) . If we know that the sampling interval - the resolution - equates to 250 microns per pixel, the estimated grain size is 8 X 250 = 2000 microns.

This information, even if it corresponds to the average grain size along the selected line, is very relevant, since the geologists often want to highlight the existence of certain trends, such as the presence of an upward fining trend. Sometimes, a large-scale downward coarsening trend 600, for instance may comprise a number of individual fining upward cycles 610, as shown in Figure 6 (reproduced from G A Blackbourn, Cores and Core Logging for Geologists, Whittles Publishing, 1990, p 63) . This indicator can be very valuable in highlighting features which might not otherwise be immediately obvious, for example it may indicate particular feature such as a weathering zone within the core.

Claims

1. A method of operating a data processing system to estimate grain size in a geological sample, comprising the steps of:

retrieving from a storage device data of the geological sample, the data storage device storing data of the geological sample which incorporates at least one characteristic varying with the grain size;

responsive to a selection of a portion of the data, retrieving a subset of the data corresponding to the selected portion, the subset of data incorporating said at least one characteristic varying with the grain size;

evaluating a set of variations of said at least one characteristic in said subset of the data; and

determining an estimate of the average grain size of the portion with respect to a dominant variation of said set.

A method as claimed in claim 1 wherein:

the step of retrieving from the storage device data of the geological sample comprises retrieving data necessary to display a digital image of the geological sample representative of said data, and displaying such an image on a display device;

the step of retrieving a subset of the data corresponding to the portion is responsive to a user selection of a portion of the image; and

the step of determining an estimate of the average grain size includes the step of making available to the user such an estimate.

3. A method as claimed in any preceding claim wherein the dominant variation is identified by applying a Fourier Transform to said subset of the data so to generate a new representation of said subset based on spatial variations.

4. A method as claimed in any preceding claim wherein the estimate is determined by dividing the dominant variation in the at least one characteristic by the resolution of the data.

5. A method as claimed in any of claims 2 to 4 wherein the user selection of the portion of the image is limited to the selection of at least one selecting parameter, the others being preset.

6. A method as claimed in claim 5 wherein the preset selecting parameters include the angle and the dimension of the portion and the at least one parameter selected by the user includes the location of the portion.

7. A method as claimed in any preceding claim wherein said portion is a line.

8. A method as claimed in any preceding claim wherein the at least one characteristic is light intensity.

9. A method as claimed in any of claims 2 to 8 wherein the displayed image also incorporates at least a second characteristic of the geological sample which provides additional information to the user to facilitate the selection of the portion to be evaluated.

10. A method as claimed in claim 9 wherein the at least second characteristic is colour.

11. A method as claimed in any of claims 2 to 10 wherein the step of making available to the user such an estimate comprises the step of displaying a slider bar indicating the estimate, the bar being annotated with a set of different grades of grain size.

12. A data processing system for estimating grain size in a geological sample comprising:

a retrieval means for retrieving data of the geological sample from a storage device, the storage device (20) storing data of the geological sample which incorporates at least one characteristic varying with the grain size;

selection means for selecting a portion of the data retrieved;

the retrieval means being responsive to the selection means to retrieve a subset of the data corresponding to the selected portion, the subset of data incorporating the at least one characteristic varying with the grain size;

evaluating means for evaluating a set of variations of said at least one characteristic in said subset of the data; and

estimating means for determining an estimate of the average grain size of the portion with respect to a dominant variation in said set.

13. A system as claimed in claim 12 further comprising:

a display means for producing a digital image of the geological sample representative of the data retrieved by the retrieval means and for displaying the digital image on a display device (38) connectable to the system; and

the selection means being responsive to a signal received from an input device (24, 26, 32) connectable to the system, to enable a user to select a portion (310) of the displayed image.